Académique Documents
Professionnel Documents
Culture Documents
CHAPTER 8
AN INTRODUCTION TO
PHARMACOGENOMICS
Ashim Malhotra
the approach to drug therapy has been largely empiric and indicate positive selection at work and that these differences in
based on clinical studies that determined the maximally toler- SNPs importantly account for the differential drugs response
ated dose and reasonable toxicity in a narrowly defined popu- in any given population (Li et al, 2011). Not only does this work
lation. This approach typically leads to the safe and effective support and explain the origin of genetic polymorphism in
administration of drugs for most individuals. However, with drug-metabolizing enzymes, it purports to provide an evolu-
empiric therapy, interindividual (allotypic) variation in drug re- tionary rationale for such differences across ethnicities.
sponse occurswith patient outcomes varying from a complete
absence of therapeutic response to potentially life-threatening GENETICS REVISITED
adverse drug reactions (ADRs).
Genetic differences may account in part for some of the An individuals genetic makeup (or genotype) is derived as a
well-documented variability in response to drug therapy. Ob- result of genetic recombination or mixing of genes from
viously, many factors other than geneticssuch as age, sex, that individuals parents. All the DNA contained in any in-
other drugs administered, and underlying disease statesalso dividual cell is known as the genome of the individual, a word
contribute to variation in drug response. However, inherited formed by the combination of gene and chromosome,
differences in the metabolism and disposition of drugs and ge- and thus represents all the genes that individual can express.
netic polymorphisms in the targets of drug therapy (e.g., me- Interestingly, even though two unrelated people share about
tabolizing enzymes or protein receptors) can have an even 99.9% of the same DNA sequences, the less than 0.1% differ-
greater influence on the efficacy and toxicity of medications. ence between them translates into a difference of 3 million
Interestingly, age, gender, and endemic geographical differ- nucleotides. These variants, introduced above, are the SNPs
ences may themselves emerge as phenotypic consequences of (pronounced snips) (Howe, 2009). The variability of the
differential epigenetic control. This implies that heterogeneity genome at these various SNPs accounts for nearly all of the
in the control of gene expression based upon age, gender, and phenotypic differences we see in each other.
geographic location is itself a life-long changing process that is The Human Genome Project has sought not only to
under the control of molecular epigenetic switches that either identify and correlate SNPs with phenotypic differences
activate or inhibit groups of genes as a unit. Specific identifi- but also to record and map haplotypes as well (Nebert,
cation of these epigenetic controls in special populations, for Zhang, & Vesell, 2008). Haplotypes are large portions of
instance differences in pediatric or geriatric protein expression genetic material (around 25,000 base pairs) that tend to
in immune cells when compared to the general adult popula- travel together. Understanding how SNPs and haplotypes
tion, can provide valuable clues to how special populations make humans genetically unique is the current focus of
based on age, gender, pregnancy, and even geographical loca- much genetic research (Nebert et al, 2008). The comple-
tion respond differentially to specific drugs. This information tion of the Human Genome Project, as well as the mapping
can then be incorporated in optimal therapeutic design. of SNPs and haplotypes, has allowed the field of pharma-
With the publication in 2001 of Landers and Venters cogenomics to understand the variability of drug metabo-
description of their groundbreaking effort to map the entire lism seen across individuals and populations. Box 8-1
human genome, the Human Genome Project (HGP), about
95% of the sequence of all human DNA was established, result-
ing in the identification of Open Reading Frames (ORF) for BOX 81 DEFINITIONS
many human proteins. A more recent development has been
the discovery of single-nucleotide polymorphisms (SNPs), Genetic polymorphism: multiple differences of a DNA
genetic differences that account for allotypic phenotype varia- sequence found in at least 1% of the population
tions. About 1.4 million SNPs in humans have been identified Genetics: the study of heredity and its variations
through a mass effort by the SNP Consortium, which was Genomics: the study of the complete set of genetic in-
formation present in a cell, an organism, or species
funded by multiple pharmaceutical companies. The existence
Pharmacogenetics: the study of the influence of hered-
of the SNP Consortium is an excellent reminder of the signifi-
itary factors on the response of individual organisms
cance of SNPs to drug companies, since SNPs may account for to drugs (Venes, 2005); the study of variations of DNA
some of the differences in drug responses seen in pharma- and RNA characteristics as related to drug response
cotherapy in the population at large (Howe, 2009). With the (U.S. Food and Drug Administration, 2010b)
identification of the individual SNPs, our understanding of Pharmacogenomics: the study of the effects of genetic
pharmacogenetics and pharmacogenomics has exploded. differences among people and the impact that these
A study published in 2011 by Li, Zhang, Zhou, Stoneking, differences have on the uptake, effectiveness, toxic-
and Tang on the heterogeneity in drug-metabolizing genes ity, and metabolism of drugs
in globally defined populations has provided profound in- SNP: single-nucleotide polymorphism
sights and ever stronger evidence for the significance and rel-
evance of SNP-induced variation in drug metabolism. This Source: Venes, D. (2005). Tabers cyclopedic medical dictionary
(21st ed.). Philadelphia: FA Davis; U.S. Food and Drug Adminis-
study compared differences in 283 drug-metabolizing en- tration. (2010b). Table of valid genomic biomarkers in the
zymes and transporter genes across 62 globally distributed context of approved drug labels. Retrieved from http://www
ethnic groups and demonstrated that patterns of emergence .fda.gov/RegulatoryInformation/Guidances/ucm129286.htm
of SNPs in specific populations spread out across the world
3827_Ch08_103-114 02/06/15 11:48 am Page 105
provides definitions of terms used in pharmacogenetics to enhance drug therapy to maximize efficacy, to target drugs
and pharmacogenomics. only to those patients who are likely to respond, and to avoid
Another interesting aspect of this discussion is the fre- ADRs. Increasing the number of patients who respond to a
quency with which the mutant gene copy is expressed. If therapeutic regimen with a concomitant decrease in the inci-
the variant copy of a gene, such as is common for genes dence of ADRs is the promise of pharmacogenomic informa-
encoding Drug Metabolizing Enzymes (DME), is expressed tion. The long-term expected benefits of pharmacogenomics
in the equivalent of 1% or more of the population, the genetic are selective and potent drugs, more accurate methods of
variation is referred to as a polymorphic variation. determining appropriate drug dosages, advanced screening
Standard adopted nomenclature is used in pharmacoge- for disease, and a decrease in the overall cost to the health-care
nomics and pharmacogenetics. Of the various mutant vari- system in the United States caused by ineffective drug therapy.
ants of a specific gene, each variant is numerically and
sequentially named starting with the wild-type or normal
or nonmutated copy of the gene. Thus, for instance, CYP2D6
GENETIC DIFFERENCES IN DRUG
written in italics refers to the normal copy of the gene, METABOLISM
whereas CYP2D6*1 (pronounced star 1) refers to the first Genetic differences in metabolism were first realized by the ob-
identified natural variant (mutant) copy of this gene. servation that sometimes very low or very high concentrations
of drug were found in some patients despite their having been
given the same amount of drug. Most genetic differences in
HISTORY OF PHARMACOGENETICS drug metabolism have been found to be monogenic genetic
The Greek philosopher and mathematician Pythagoras polymorphisms, meaning that they arise from the variation in
recorded the first interindividual difference of drug admin- one gene (Nebert et al, 2008).
istration in 510 BCE when he noted that some patients de-
veloped hemolytic anemia after ingesting the fava bean Genetic Polymorphism
(Nebert et al, 2008). The term pharmacogenetics was first
coined by Vogel in 1959, but not until1962 was pharmaco- A genetic polymorphism occurs when a difference in the
genetics defined as the study of heredity and the response to allele(s) responsible for the variation is a common occurrence.
drugs by Kalow (Nebert et al, 2008). Since 1962, the term has An allele is an alternative form of a gene. A gene is called poly-
been used to refer to the effects of genetic differences on a morphic when allelic variations occur throughout a given pop-
persons response to drugs. ulation at a stable rate of less than 1% (Howe, 2009). Under such
Interest in pharmacogenetics emerged in the 1950s in re- circumstances, mutant genes will exist somewhat frequently
sponse to the discovery of an abnormal butyrylcholinesterase alongside wild-type genes. The mutant genes will encode for
enzyme in psychiatric patients who exhibited prolonged mus- the production of mutant proteins in these populations. The
cular paralysis after administration of succinylcholine before mutant proteins will, in turn, interact with drugs in different
electroconvulsive therapy (Meyer, 2004). Also in the 1950s a manners, sometimes slight, sometimes significant. Monogenic
traits by themselves cannot explain the complexity of drug me-
connection was established between the development of hemol-
tabolism (Nebert et al, 2008). Genes interact on a complex level,
ysis in African American males treated for malaria with pri-
yielding different responses depending on which genes are wild-
maquine and glucose-6-phosphate dehydrogenase deficiency
type and which show mutant phenotypes. Sometimes these
(Beutler, 1959). Other seminal pharmacogenetic findings in-
interactions can be very difficult to elucidate and may in fact
clude the identification of the proportion of slow acetylators in
be the source of seemingly unexplainable drug reactions.
certain ethnic groups, including 10% of the Japanese and Eski-
mos; 20% of the Chinese; and 60% Caucasians, blacks, and Figure 8-1 illustrates the relationship between genetic polymor-
South Indians (Ellard, 1976), and attribution of peripheral neu- phisms in drug metabolism and at drug receptors.
ropathy to slow acetylation of isoniazid in some patients treated Four different phenotypes categorize the effects that ge-
for tuberculosis due to genetic diversity in the enzyme N- netic polymorphisms have on individuals: poor metabolizers
acetyltransferase (Fig. 8-1) (Yamamoto, Subue, Mukoyama, (PMs) lack a working enzyme; intermediate metabolizers
Matsuoka, & Mitsuma, 1999). The rate of acetylation of a drugs (IMs) are heterogeneous for one working, wild-type allele
such as isoniazid is clinically relevant because it determines the and one mutant allele (or two reduced-function alleles); ex-
tensive metabolizers (EMs) have two normally functioning
rate of elimination of the drug from the body. Thus, individuals
alleles; and ultrarapid metabolizers (UMs) have more than
known as slow acetylators will metabolize the drug slowly, al-
one functioning copy of a certain enzyme (Belle & Singh,
lowing greater residence time in the body and enhanced toxic-
2008). See Table 8-1 for the clinical implications of genetic
ity. It is pharmacogenomic variation which is responsible for
polymorphisms.
slow or fast acetylators as explained below.
No. subjects
Efficacy
Toxicity 12 Slow rate of acetylation
100 100
Drug Conc.
wt/wt wt/wt 75 1
Effect (%)
50 50 wt/m 35 1
30 m/m 10 1
0 0
0 24 h 0 50 100
A
0
100 100
0 4 8 12
Drug Conc.
50 50 wt/m
polymorphism.
50 >80
99
m/m 10 >80
0 0
C 0 24 h 0 50 100 Phase I metabolism enzymes are responsible for approxi-
Time Drug Concentration
mately 59% of the adverse drug reactions cited in the litera-
Figure 81. Genetic polymorphisms and drug metabolism/receptors. ture (Phillips, Veenstra, Oren, Lee, & Sadee, 2001). In terms
of evolution, the cytochrome P450 (CYP450) enzyme system
was one of the first biocatalytic machineries to emerge on
Table 81 Clinical Implications of Genetic earth. These ubiquitous enzymes contain an iron-porphyrin
Polymorphisms ring center that is essential to the chemical reaction they cat-
alyze. During this oxygenation reaction, the oxidative state
Metabolizer Effect on Clinical of iron in the porphyrin ring changes, resulting in spec-
Phenotype Drug Metabolism Implications
trophotometric absorption maxima observed at 450 nm,
Poor to Slow Prodrug will be metabolized which contributed to their naming.
intermediate slowly into active drug CYPs are generally located in the endoplasmic reticulum
metabolizers metabolite. May have accu- (ER) and the mitochondria in human cells, of which the ER iso-
mulation of prodrug.
forms are of particular importance to the field of drug metabo-
Active drug will be metabo-
lized slowly into inactive lism. In terms of their organ distribution, they are found in
metabolite. Potential for ac- greater amounts in the liver and the intestine and to a somewhat
cumulation of active drug. lesser extent in other organs, such as the skin, brain, lungs, and
Patient requires lower kidneys. Hepatic, renal, and intestinal ER CYPs are involved in
dosage of medication.
the biotransformation of a plethora of drugs and endogenous
Ultrarapid Fast Prodrug rapidly metabolized substrates in humans mainly by oxygenation of the target sub-
metabolizers into active drug. No strate molecule and mediated by differential oxidation states of
dosage adjustment needed. the central iron atom in the enzyme. Due to this oxygenation
Active drug rapidly metabo-
lized into inactive metabo-
reaction, CYPs are classified as monooxygenases. The high ge-
lites leading to potential netic variability of the cytochrome P450 enzymes constitutes
therapeutic failure. Patient the most important of the phase I metabolizing enzymes, with
requires higher dosage of a total of 57 genes encoding for CYP450 enzymes. Of these,
active drug. CYP2D6, CYP2C9, and CYP2C19 are highly polymorphic and
account for upward of about 40% of hepatic phase I metabolism
(Phillips et al, 2001) (Fig. 8-3 and Table 8-2).
mostly in the liver and is divided into two major categories,
phase I (oxidation, reduction, and hydrolysis reactions)
Specific CYP450 Enzymes
and phase II metabolism (conjugation reactions). A hall-
mark experiment in pharmacogenomics, diagrammed in CYP2D6
Figure 8-2, illustrates how differences in the rates of the Up to 25% of drugs are metabolized via CYP2D6 (Belle &
phase II metabolizing enzyme N-acetyltransferase (NAT-2) Singh, 2008). Phenotypic variations between some enzymes
can affect the half-life and plasma concentration of drugs can have an astounding outcome on drug therapy. For exam-
that are subject to NAT-2 metabolism (Meyer, 2004). ple, a 1,000-fold difference has been observed in the rate of
3827_Ch08_103-114 02/06/15 11:48 am Page 107
Drug-Metabolizing Enzymes
CYP2C9 Tolbutamide, warfarin, phenytoin, NSAIDS Anticoagulant effect of warfarin
CYP2D6 Beta blockers, antidepressants, antipsychotics, Tardive dyskinesia from antipsychotics; narcotic side
codeine, debrisoquine, dextromethorphan, effects, efficacy, and dependence: imipramine dose
encainide, flecainide, guanoxan, methoxyam- requirement; beta blocker
phetamine, N-propylamaline, perhexiline, effect
phenacetin, phenformin, propafenone,
sparteine
Thiopurine methyltransferase Mercaptopurine, thioguanine, azathioprine Thiopurine toxicity and efficacy; risk of second cancers
Drug Targets
ACE Enalapril, lisinopril, captopril Renoprotective effects, cardiac indices, blood pressure,
immunoglobulin A nephropathy
30 29
25
Percentage of individuals
20
20
15
10
10
Figure 85. Percentage distribution of individ-
uals across countries showing a duplication of
an allele of CYP2D6. The figure explains the 5
3.7
exaggerated metabolism of some drugs in the
2
specified percentage of individuals belonging
to certain ethnic backgrounds (generously as- 0
suming ethnic homogeneity in some coun- Sweden Germany Spain Saudi Arabia Ethiopia
tries) due to increased 2D6 activity.
Haining, Bajpai, & Levy, 1999), resulting in adjusted dose maintenance of anticoagulant therapy following an adverse
requirements in these patients compared with noncarriers. cardiovascular event, in part due to some conflicting reports
In addition, patients with homozygous presentation of a in the literature about the clinical effectiveness and relevance
CYP2C9 mutation appear to have a greater reduction in of pharmacogenomic variability in warfarin drug metabolism.
dosing requirement than do heterozygotes. Approximately However, recent studies on the pharmacogenomics of warfarin
one-third of the population carries at least one allele for the have included an analysis of the effects of polymorphism of
slow-metabolizing form of CYP2C9 (U.S. Food and Drug CYP2C9 and vitamin K epoxide reductase (VKORC1) across
Administration, 2009). The clinical implications of altered populations in Europe, Asia, and elsewhere. Polymorphisms
warfarin metabolism can be significant; the clinical implica- in these two genes account for nearly 40% of the differences in
tions of pharmacogenomic variants are found later in this warfarin therapy across populations (Yip & Pirmohamed,
chapter. See Table 8-4. 2013). Taken together, the data from these pharmacogenomic
In spite of the information outlined above, controversy ex- studies indicate a strong connection between the variant effects
ists in the field of pharmacogenetic testing over initiation or of CYP2C9 polymorphism and the metabolism and therapeu-
tic efficacy of warfarin and acenocoumaral. CYP2C9 variation
was reported to be important for the maintenance therapy of
Table 84 CYP2C (9 and 19)
warfarin in a genome-wide association analysis in the Swedish
Substrate Inhibitors Inducers and Japanese population (Cha et al, 2010).
S-warfarin Amiodarone Carbamazepine
Data from disparate research resources need to be orga-
nized to obtain clinically relevant information that assists in
Losartan Cimetidine Phenytoin guiding the therapeutic rationale for the use of warfarin in
Diazepam Chloramphenicol Rifampin special populations. One such approach would be to analyze
studies on the frequency of allelic variation of the three drug-
Imipramine Fluconazole
metabolizing genes for warfarin in selected populations. Thus,
Amitriptyline Isoniazid for the CYP2C9 gene, the frequency of allelic variation for
CYP2C9*2 is about 10% in Caucasians, compared to less than
Phenytoin Ketoconazole
1% in Africans and Asians, while the frequency of allelic vari-
Rosiglitazone Zafirlukast ation for CYP2C9*3 is about 6% in Caucasians, 4% in Asians,
Fluoxetine and less than 1% in Africans (Fig. 8-6). The frequency of allelic
variation of the VKORC1 gene is substantially higher in all
Fluvoxamine the three ethnicities compared with the mutational frequency
Sertraline of CYP2C9. For the specific VKORC1 (-1639) mutation, al-
lelic frequency is 98% for Africans, 60% for Caucasians, and
Rosiglitazone
2% for Asians; see Figure 8-7 (Voora & Ginsburg, 2012).
3827_Ch08_103-114 02/06/15 11:48 am Page 110
CYP3A4
6
The CYP3A group of isoenzymes is responsible for up to
50% of drug metabolism (Howe, 2009). CYP3A4 isoen-
4 zyme is responsible for metabolism of several important
classes of drugs that are commonly used in primary care
(see Table 8-1). Examples of these classes include azole an-
2 tifungals, calcium channel blockers, antihistamines, anti-
convulsants, antimicrobials, and corticosteroids. Both
drug-related induction or inhibition of CYP450 3A4 isoen-
0
CYP2C9*2 CYP2C9*3 zyme may complicate drug therapy in patients (Howe,
CYP2C9 Allelic Mutations 2009). Predicting the onset and offset of these effects is very
difficult. The time to onset and offset of drugdrug inter-
Figure 86. Percentage distribution of individuals across ethnicities actions is closely related to each drugs half-life and the
exhibiting polymorphism in CYP2C9. half-life of enzyme production. Clinically significant drug
interactions in this setting may increase the risk of toxicity.
For example, amiodarone has a half-life of close to 60 days
Percentage of Allelic Variation VKORC1 and requires months to reach steady state and inhibit the
Across Different Ethnicities
CYP450 enzyme system effectively (Table 8-5). Conversely,
Caucasians Africans Asians
120
Table 85 CYP3A4
100
Substrate Inhibitors Inducers
Percentage of Allelic Frequency
Sildenafil Fluoxetine
Figure 87. Percentage distribution of individuals across ethnicities
showing variation in VKORC1. Astemizole Zileuton, zafirlukast
Terfenadine Verapamil
The wealth of scientific data documenting evidence of
CYP2C9*2, CYP2C9*3, and VKORC1 (-1639 G > A) poly- Pioglitazone Amiodarone
morphisms affecting dose response to warfarin in different R-warfarin Corticosteroids
populations resulted in the FDA updating the label of the
Fluvoxamine
drug in 2007 and then again in 2010 (Finkelman et al,
3827_Ch08_103-114 02/06/15 11:48 am Page 111
it takes less than 2 days for rifampin, which is a nonspecific Small Intestine Intestinal Wall Enteric Blood Flow
CYP450 inducer with a shorter half-life, to decrease blood CYP3A4
concentrations of many drugs to a subtherapeutic level and
significantly increase the risk of therapeutic failure. Close
Metabolite
monitoring is required when prescribing drugs that induce
or inhibit CYP3A4 enzymes. See Table 8-5 for further
information. Parent
Drug
P-Glycoprotein
P-GLYCOPROTEIN
P-glycoprotein (Pgp) is a membrane-bound, ATP-depen-
Figure 88. Drugmetabolism interactions.
dent transport system responsible for the efflux of a variety
of xenobiotics from cells to the extracellular fluid. This in-
cludes the ejection of drugs from cells, usually against their
concentration gradients. Pgp, also known as multidrug re-
CLINICAL IMPLICATIONS
sistance (MDR1) protein, is the product of the ABCB1 and OF PHARMACOGENOMICS
ABCB4 genes and is a member of adenosine triphosphate
(ATP)-binding family of proteins. Differential expression Adverse Drug Reactions
of Pgp may explain tissue-specific and temporal variations One benefit of understanding pharmacogenomics is the pos-
in efflux efficiency in different cells. In fact, chemoresis- sibility of a decrease in the number of ADRs. The CYPP450
tance in cancer therapeutics has been strongly linked with enzymes in families 1 to 3 mediate 78% to 80% of all phase I
Pgp expressionthe more Pgp protein expressed by the dependent metabolism of clinically used drugs (Spatzenegger
cell, the greater the efflux potential of xenobiotics such as & Jaeger, 1995). The polymorphic forms of CYPP450s are
anticancer drugs. In addition to differences in protein ex- responsible for the development of idiosyncratic ADRs
pression, polymorphic variation of the Pgp genes may also (Kalgutkar, Obach, & Maurer, 2007). According to Phillips
dynamically affect intracellular and plasma drug concen- and Van Bebber (2005), 56% of drugs cited in ADR studies
tration. Over 50 SNPs within the ABCB1 gene have been are metabolized by polymorphic phase I enzymes, of which
identified, which may lead to variability in drug responses 86% are P450s.
(Reed & Parissenti, 2011).
Pharmaceutically relevant examples of this include the
variation in drug response to agents such as antiepileptic Warfarin
drugs, select cardiovascular agents, and so on. Interestingly, In 2008, the package insert for warfarin was updated by the
P-glycoprotein at the site of the gastrointestinal (GI) tract ef- U.S. Food and Drug Administration to include application
fluxes hydrophilic drugs out of the cell and inhibits drug ab- of pharmacogenomics to the dosing of warfarin. Previous
sorption through the GI tract (Howe, 2009). As drugs work had identified variable metabolism by CYP2C9 as a
passively diffuse through the GI tract, Pgp pumps move drugs major contributor to the variable response to the drug. In
from cytoplasmic areas to extracellular fluid. Some examples 2004, coding-region mutations in VCORC1, encoding a sub-
of substrates of P-glycoprotein include carvedilol, diltiazem, unit of the vitamin K epoxide reductase complex (the phar-
and digoxin (Howe, 2009). Several antiepileptic drugs such macological target for the drug), were found to cause a rare
as phenytoin, carbamezapine, lamotrizine, phenobarbital, syndrome of warfarin resistance. Subsequently, the variants
valproic acid, and gabapentin are substrates or inhibitors of in VCORC1 have been found to account for a much greater
Pgp, but there is considerable controversy in the literature re- fraction of variability in warfarin response (21%) than do
garding these. In the case of digoxin, Pgp affects the level of variations in CYP2C9 (6%) (Gulseth et al, 2009). Although
digoxin available for absorption and elimination (Howe, genetic testing prior to prescribing has not yet been required
2009). P-glycoprotein inhibitors include verapamil, quini- by the FDA, numerous warfarin dosing calculators exist on
dine, cyclosporine, and ketoconazole (Howe, 2009). If an in- the Web where a clinician can insert clinical information
hibitor of P-glycoprotein is administered, then blood levels about the patient, including genetic test results and indica-
of substrates will rise, as seen if quinidine is administered tions, and a dosing regimen will be calculated or individu-
with digoxin. alized for that patient (http://www.warfarindosing.com;
Drugs can be categorized as reversible or suicidal in- http://www.globalrph .com/warfarin.htm).
hibitors or P-glycoproteins. For example, calcium channel
blockers and high-dose steroids are considered as re-
versible inhibitors of both P-glycoproteins and CYP450.
Pharmacogenetic Testing Prior
However, grapefruit and ritonavir are suicidal agents for to Prescribing
both P-glycoprotein and CYP450, meaning the effect of The FDA now requires additional pharmacogenomic in-
grapefruit juice will be prolonged, perhaps up to 24 hours. formation on several drug package inserts (Table 8-6).
See Figure 8-8. Within the anticoagulant drug class, warfarin is a drug with
3827_Ch08_103-114 02/06/15 11:48 am Page 112
Table 86 U.S. Food and Drug Administration a narrow therapeutic index. In 2007, the FDA recom-
Positions on Necessity of mended an important label update suggesting genetic test-
Pharmacogenetic Testing as ing to prevent possibly fatal bleeding in patients with
Indicated on Drug Labeling polymorphic variants of CYP2C9, the metabolizing enzyme,
and VKORC1, the target enzyme of warfarin. Patients with
Pharmacogenetic
Biomarker Drug CYP2C9 variations require more time to achieve the Inter-
national Normalized Ratio, or INR, and are at an increased
Test Required risk for bleeding (Sconce, 2005); they may also require lower
EGFR expression Cetuximab doses of warfarin to achieve and maintain therapeutic INR
HER2/NEU overexpression Trastuzumab (Limdi, 2007). Thus, if there are indications of inherited dif-
ferences in these genes, the patient should be genotyped.
CCR-5-tropic HIV-1 Maraviroc
However, monitoring INR is still as much of a requirement
Presence of Philadelphia Dasatinibchromosome while dosing warfarin as before. While the FDA did not ex-
plicitly require genetic testing in patients prior to prescrib-
Test Recommended
ing warfarin, the package labeling did show changes in
HLA-B*1502 Carbamazepine
dosage amounts. There is an FDA-approved genetic testing
HLA-B*5701 Abacavir kit available, but others may also be used. Generally, cell
CYP2C9 variants Warfarin samples are collected from the mouth or from blood. How-
ever, it should be emphasized that genetic testing is not the
VKORC1 variants Warfarin sole consideration, since patient-related factors such as age,
Protein C deficiency Warfarin sex, body weight, and some other parameters may need to
be considered with the genetic results. A variety of online
TPMT variants Azathioprine, mercaptopurine,
thioguanine
algorithms can aid physicians and hospital staff in making
these dosage adjustments, as presented, for example, at
UGT1A1 variants Irinotecan www.warfarindosing.org.
G6PD deficiency Rasburicase The pharmacogenetic tests mentioned on drug labels
can be classified as test required, test recommended, and
Urea cycle disorders Valproic acid
information only. Currently, four drugs are required to
Information Only have pharmacogenetic testing performed before they are pre-
c-KIT expression Imatinib scribed: cetuximab, trastuzumab, maraviroc, and dasatinib.
Cetuximab treatment needs a confirmation of epidermal
CYP2C19 variants Voriconazole
growth factor receptor (EGFR) expression. Trastuzumab
CYP2C9 variants Celecoxib therapy requires testing for HER2/NEU overexpression.
CYP2D6 variants Atomoxetine, tamoxifen, Infection with CCR-5-tropic HIV-1 should be confirmed be-
fluoxetine fore initiation of therapy with maraviroc (an antiretroviral).
Dasatinib is used for the treatment of patients with Philadelphia
DPD deficiency Capecitabine, fluorouracil
chromosome-positive acute lymphoblastic leukemia resistant
EGFR expression Erlotinib to or intolerant of prior therapy (U.S. Food and Drug Ad-
G6PD deficiency Rasburicase, primaquine
ministration, 2010b).
In December 2007, the FDA added a Black-Box Warning
NAT variants Isoniazid, rifampin on the carbamazepine label, recommending testing for the
Philadelphia chromosome Busulfandeficiency HLA-B*1502 allele in patients with Asian ancestry before ini-
tiating carbamazepine therapy because these patients are at
PML/RAR gene expression Tretinoin
high risk of developing carbamazepine-induced Stevens
EGFR = epidermal growth factor receptor; HER2/NEU = v-erb-b2 erythro- Johnson syndrome (SJS) or toxic epidermal necrolysis (TEN).
blastic leukemia viral oncogene homolog 2; CCR-5 = chemokine C-C motif Interestingly, although Asians or patients with Asian ancestry
receptor; HLA = human leukocyte antigen; CYP2C9 = cytochrome P-450 have been reported to have a strikingly high frequency (10
isoenzyme 2C9; VKORC1 = vitamin K epoxide reductase complex subunit 1;
TPMT = thiopurine S-methyltransferase; UGT1A1 = uridine diphosphate-
times higher than whites) of carbamazepine-induced SJS or
glucuronosyltransferase 1A1; c-KIT = v-kit Hardy-Zuckerman 4 feline TEN if they carry an HLA-B*1502 allele, other races carrying
sarcoma viral oncogene homolog; CYP2C19 = cytochrome P-450 2C19; the allele do not seem to have the increased risk (U.S. Food
CYP2D6 = cytochrome P-450 isoenzyme 2D6; DPD deficiency = dihydropy- and Drug Administration, 2007).
rimidine dehydrogenase; G6PD = glucose-6-phosphate dehydrogenase; NAT The anticancer agent irinotecan is a prodrug used for
= N-acetyltransferase; PML/RAR = promyelocytic leukemia/retinoic acid
receptor.cytochrome P-450 isoenzyme 2D6; DPD deficiency = dihydropy-
the treatment of colorectal cancer, small-cell lung cancer,
rimidine dehydrogenase; G6PD = glucose-6-phosphate dehydrogenase. and other solid tumors. The active metabolite of irinotecan
Source: Derived from U.S. Food and Drug Administration. (2010b). Table is SN-38, a topoisomerase I inhibitor, and uridine diphos-
of valid genomic biomarkers in the context of approved drug labels. phate glucuronosyltransferase 1A1 (UGT1A1) plays a crit-
Retrieved from http://www.fda.gov/Drugs/ScienceResearch/ResearchAreas/ ical role in inactivating SN-38 (McLeod & Hoskins, 2007).
Pharmacogenetics/ucm083378.htm
3827_Ch08_103-114 02/06/15 11:48 am Page 113
The low activity of the UGT1A1 enzyme may increase the REFERENCES
risk for adverse events associated with irinotecan therapy
Arias, E. (2010). United States life tables, 2006. National Vital Statistics Re-
(e.g., neutropenia) by increasing serum concentrations of ports, 58(21). Retrieved from http://www.cdc.gov/nchs/data/ nvsr/nvsr58/
the active metabolite. A polymorphism in the promoter nvsr58_21.pdf
region of the UGT1A1 gene determines patient exposure Belle, D. J., & Singh, H. (2008). Genetic factors in drug metabolism. American
to SN-38 (McLeod & Hoskins, 2007). Patients homozy- Family Physician, 77(11), 15531560.
gous for the polymorphism (UGT1A1*28) are at a 5-fold Beutler, E. (1959). The hemolytic effect of primaquine and related compounds:
A review. Blood, 14(2), 103139.
greater risk of irinotecan-related toxicity compared with Cha, P. C, Mushiroda,T., Takahashi, A., Kubo, M., Minami, S., Kamatani,
patients with one or two normal alleles. Additionally, the N., et al. (2010). Genome-wide association study identifies genetic deter-
FDA has approved a test for detection of the UGT1A1*28 minants of warfarin responsiveness for Japanese. Human Molecular
genotype for irinotecan dosing. Additional genotype tests Genetics, 19, 47354744.
approved by the FDA and their implications are summa- Ellard, G. A. (1976). Variations between individuals and populations in the
acetylation of isoniazid and its significance for the treatment of pulmonary
rized in Table 8-7. tuberculosis. Clinical Pharmacology and Therapeutics, 19(5, Pt. 2), 610625.
Evans, W. E., & Relling, M. V. (1999). Pharmacogenomics: Translating
functional genomics into rational therapeutics. Science, 286, 487491.
SUMMARY Finkelman, B. S., Gage, B. F., Johnson, J. A., et al. (2011). Genetic warfarin
dosing: Tables versus algorithms. Journal of the American College of
We live in remarkable times, in which multiple therapeutic Cardiology, 57, 612618.
options are available for most common diseases. However, Gulseth, M. P., Grice, G. R., & Dager, W. E. (2009). Pharmacogenomics of
the selection of the optimal medication for an individual warfarin: Uncovering a piece of the warfarin mystery. American Journal
patient is still problematic. Practitioners still pick the of Health Systems Pharmacists, 66, 123133.
right initial medication only half the time and ADRs are Howe, L. A. (2009). Pharmacogenomics and management of cardiovascular
disease. The Nurse Practitioner, 34(8), 2835.
still unpredictable. In addition, the expense of new biolog- Kalgutkar, A. S., Obach, R. S., & Maurer, T. S. (2007). Mechanism-based in-
ical agents is such that even wealthy countries like the activation of cytochrome P450 enzymes: Chemical mechanisms, structure-
United States cannot afford to treat all patients. activity relationships and relationship to clinical drug-drug interactions
The completion of the Human Genome Project has en- and idiosyncratic adverse drugs reactions. Current Drug Metabolism, 8,
abled the development of clinical tools for patient evaluation. 407447.
Kolata G. In treatment of leukemia, glimpses of the future. New York Times.
Pharmacogenomics may allow identification of patients July 7, 2012.
most likely to benefit from a given therapy and those patients Li, J., Zhang, L., Zhou, H., Stoneking, M., & Tang, K. (2011). Global patterns
for whom the cost and risk outweigh the benefits. Both the of genetic diversity and signals of natural selection for human ADME
safety and efficacy of drug therapy may improve. In the fu- genes. Human Molecular Genetics, 20(3), 528540.
ture, genotyping may be used to personalize drug treatment McLeod, H. L., & Hoskins. J. M. (2007). Personalized drug therapy: The era
of pharmacoeconomics. Retrieved from http://www.ipit.unc.edu/ files/
for vast numbers of patients, decreasing the cost of drug 1233076125-one.pdf
treatment and increasing the efficacy of drugs and health in Meyer, U. A. (2004). Pharmacogeneticsfive decades of therapeutic lessons
general. from genetic diversity. Nature Reviews, Genetics, 5(9), 669676.
Nebert, D. W., Zhang, G., & Vesell, E. S. (2008). From human genetics and
genomics to pharmacogenetics and pharmacogenomics: Past lessons,
future directions. Drug Metabolism Reviews, 40, 187224.
Table 87 FDA-Approved Diagnostic Test Penny, M. A., & McHale, D. (2005). Pharmacogenomics and the drug
Commercially Available for Commonly discovery pipeline: When should it be implemented? American Journal of
Prescribed Pharmacologic Therapies Pharmacogenomics, 5(1), 5362. Review.
Peters, G. J., Smorenburg, C. H., & Van Groeningen, C. J. (2004). Prospective
Genetic Test Drug Benefit of Genetic Test clinical trials using a pharmacogenetic/pharmacogenomic approach. Journal
of Chemotherapy, 16 (Suppl. 4), 2, 530.
CYP2C9/ Warfarin Reduce time to target INR;
Phillips, K. A., & Van Bebber, S. L. (2005). Measuring the value of pharma-
VCORC1 possibly decrease bleeding
cogenomics. Nature Reviews Drug Discovery, 4(6), 500509.
episodes
Phillips, K. A., Veenstra, D. L., Oren, E., Lee, J. K., & Sadee, W. (2001). Potential
CYP2D6 Tamoxifen Reduce therapeutic failure role of pharmacogenomics in reducing adverse drug reactions: A systematic
Codeine Reduce GI toxicities/ improve review. Journal of the American Medical Association, 286(18), 22702279.
pain control Rettie, A. E., Haining, R. L., Bajpai, M., & Levy, R. H. (1999). A common genetic
Oxycodone Reduce GI toxicities/ improve basis for idiosyncratic toxicity of warfarin and phenytoin. Epilepsy Research,
pain control 35, 253.
Tricyclic Reduce therapeutic failure Schroth, W., Goetz, M. P, Hamann, U., Fasching, P. A, Schmidt, M., Winter,
antidepressants S., et al. (2009). Association between CYP2D6 polymorphisms and out-
comes among women with early stage breast cancer treated with tamoxifen.
TMPT Azathiaprine Reduce myelosuppression Journal of the American Medical Association, 302, 14291436.
6-mecaptopurine Reduce myelosuppression Schroth, W., Hamann, U., Fasching, P. A., Dauser, S., Winter, S., Eichelbaum,
M., et al. (2010). CYP2D6 polymorphisms as predictors of outcome in breast
UGT1A1 Irinotecan Reduce neutropenia cancer patients treated with tamoxifen: Expanded polymorphism coverage
Source: Derived from McLeod, H. L., & Hoskins. J. M. (2007). Personalized improves risk stratification. Clinical Cancer Research, 16, 44684477.
drug therapy: The era of pharmacoeconomics. Retrieved from http://www Sistonen, J., Madadi, P., Ross, C. J., Yazdanpanah, M., Lee, J. W., Landsmeer,
.ipit.unc.edu/files/1233076125-one.pdf M. L., et al. (2012). Prediction of codeine toxicity in infants and their
3827_Ch08_103-114 02/06/15 11:48 am Page 114
mothers using a novel combination of maternal genetic markers. Clinical sample coding categories. Retrieved from http://www.fda.gov/Regulatory-
Pharmacology and Therapeutics, 91, 692699. Information/Guidances/ucm129286.htm
Spatzenegger, M., & Jaeger, W. (1995). Clinical importance of hepatic cy- U.S. Food and Drug Administration. (2010b). Table of valid genomic biomark-
tochrome P450 in drug metabolism. Drug Metabolism Reviews, 27, 397417. ers in the context of approved drug labels. Retrieved from www.fda.gov/
Suarez-Kurtz, G. (2005). Pharmacogenomics in admixed populations. Drugs/ScienceResearch/ResearchAreas/Pharmacogenetics/ucm083378.htm
Trends in Pharmacology Science, 26(4), 196201. Vermes, A., & Vermes, I. (2004). Genetic polymorphisms in cytochrome
Thompson, A. .M., Johnson, A., Quinlan, P., Hillman, G., Fontecha, M., P450 enzymes. Effect on efficacy and tolerability of HMG0CoA Reductase
Bray, S. E., et al. (2011). Comprehensive CYP2D6 genotype and adherence inhibitors. American Journal of Cardiovascular Drugs, 4(4), 247255.
affect outcome in breast cancer patients treated with tamoxifen monother- Voora, D., & Ginsburg, G. S., (2012). Clinical applications of cardiovascular
apy. Breast Cancer Research and Treatment, 125, 279287. pharmacogenetics. Journal of the American College of Cardiology, 60(1),
U.S. Department of Health, Education, and Welfare. (1974). Vital statistics of 920.
the United States, 1970 (Vol. 2, Sec. 5) (DHEW Publication No. HRA 74- Walgren, R. A., Meucci, M. A., & McLeod, H. L. (2005). Pharmacogenomic
1104). Retrieved from http://www.cdc.gov/nchs/data/lifetables/ life70.pdf discovery approaches: Will the real genes please stand up? Journal of Clinical
U.S. Food and Drug Administration. (2007). Information on carbamazepine Oncology, 23(29), 73427349.
(marketed as Carbatrol, Equetro, Tegretol and generics) with FDA alerts. Yamamoto, M., Subue, G., Mukoyama, M., Matsuoka, Y., & Mitsuma, T.
Retrieved from http://www.fda.gov/Drugs/DrugSafety (1999). Demonstration of slow acetylator genotype of N-acetyltransferase in
U.S. Food and Drug Administration. (2009). Critical path initiativewarfarin isoniazid neuropathy using an archival hematoxylin and eosin section of a
dosing. Retrieved from http://www.fda.gov/ScienceResearch/Special- sural nerve biopsy specimen. Journal of Neurological Sciences, 135(1), 5154.
Topics/CriticalPathInitiative/FacesBehindCriticalPath/ ucm077473.htm Yip, V. L., & Pirmohamed, M. (2013). Expanding role of pharmacogenomics
U.S. Food and Drug Administration. (2010a). E15 definitions for genomic in the management of cardiovascular disorders. American Journal of Car-
biomarkers, pharmacogenomics, pharmacogenetics, genomic data and diovascular Drugs, 13, 151162.